Our lab’s head, Christian Katzenbach, was a guest on Radio Bremen’s morning show on “Bremen Zwei” on Tuesday, 15 September 2026, where he spoke as an expert about the new wave of “AI will destroy us all” warnings and the resulting calls for AI regulation.

The conversation centred on the dramatic warnings issued by some AI companies about the unpredictable consequences of their own products, and on the resulting calls for regulation: is AI slipping out of control, and should we be afraid for our safety?

Katzenbach argued that the debate is indeed about responsibility and control, but not about AI achieving “world domination.” AI systems have no consciousness and no intentions of their own; they are deployed by organisations and people, who grant them resources and access rights. That is precisely why we should be more cautious: AI systems must not be handed complete control over critical infrastructure, let alone nuclear weapons. Above all, tech companies need to be held accountable. So far, they have been remarkably successful at selling the idea that ever-further AI development is both necessary and desirable, partly by invoking the argument that competitors such as China must not be allowed to get ahead. But the real danger, Katzenbach stressed, is not losing an AI race against China. It is that, in all the excitement, we, including the tech companies themselves, forget to hold ourselves to existing laws and basic principles of responsible conduct.